Cognitive Hexagon-Controlled Intelligent Speech Interaction System

摘要

Several intelligent speech interaction (ISI) systems have emerged over the past four decades that have served the human community. The research papers show that these systems are very well connected to the cognitive hexagon and the six hybrid approaches. Where this hexagon reveals six distinct cognitive areas, one of the six hybrid perspectives gives rise to the dimensions of speech quality. This survey has been undertaken to reveal the dimensions of speech quality and to discuss the role of cognitive hexagonal regions on these dimensions with hybrid approaches. Here, ISI systems support this discussion and follow them as cognitive machines. An overview of the state of the art related to ISI systems is also described here. Techniques, such as processing [natural language (NL)], speech synthesis [speech-to-text (STT) or text-to-speech (TTS)], computing (voice/mobile), and audio mining are presented in this overview. These are contributing well with technologies, such as the Internet of Things (IoT), Voice over Internet Protocol (VoIP), and cloud-based systems (CBSs). In addition, stochastic components, such as reliability, availability, and failure rate were discussed to analyze whether the Quality of Service (QoS) of these ISI systems is described. Additionally, after the discussion, some aspects of the applications are also discussed along with the essential advantages and significant drawbacks.